Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Combining nearest neighbor classifiers versus cross-validation selection.

Minhui Paik1, Yuhong Yang

  • 1Iowa State University, USA. minhui@iastate.edu

Statistical Applications in Genetics and Molecular Biology
|May 2, 2006
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Tunable Redox Dynamics of Organic Electrochemical Transistors for High-Performance Parallel Reservoir Computing.

ACS applied materials & interfaces·2026
Same author

Multi-omics and biochemical analyses provide new insights into the effects of poly-β-hydroxybutyrate on flesh quality in common carp (Cyprinus carpio).

Food chemistry·2026
Same author

Fast Click Crosslinking Bio-Inspired Adhesives With Tailored Hierarchical High Bonding and Exceptional Versatility.

Small (Weinheim an der Bergstrasse, Germany)·2026
Same author

Insights Into the Toxic Effects of 3-Monochloropropane-1,2-Diol on Sertoli Cells via Tandem Mass Tagging Proteomics.

Journal of applied toxicology : JAT·2026
Same author

Reprogrammed Propionate Metabolism Alters Redox-Dependent Aldosterone Production.

Hypertension (Dallas, Tex. : 1979)·2026
Same author

Methyl thiobutyrate: A microbial volatile compound with dual modes-of-action against root-knot nematodes.

Pest management science·2026

This study introduces a novel weighting method to improve tumor classification using gene expression data. The new approach outperforms traditional cross-validation when classifier selection is uncertain, enhancing diagnostic accuracy.

Area of Science:

  • Bioinformatics
  • Computational Biology
  • Genomics

Background:

  • Gene expression profiling is crucial for tumor classification.
  • Nearest Neighbor (NN) methods are commonly used but can be sensitive to parameter selection.
  • Cross-validation (CV) is typically employed for parameter optimization in NN classifiers.

Purpose of the Study:

  • To address the limitations of CV in NN-based tumor classification.
  • To propose and evaluate a novel weighting method for combining multiple NN classifiers.
  • To enhance the robustness and accuracy of tumor classification from gene expression data.

Main Methods:

  • Application of various discriminant methods for tumor classification.
  • Utilizing the Nearest Neighbor (NN) algorithm.

Related Experiment Videos

  • Implementing a proposed weighting method to combine multiple NN rules.
  • Comparative analysis against traditional Cross-Validation (CV) methods using four gene expression datasets.

    • Main Results:

    • Cross-validation (CV) can exhibit poor performance due to uncertainty in selecting optimal classifier parameters.
    • The proposed weighting method effectively combines multiple NN rules.
    • The combined classifier demonstrates significantly improved performance compared to CV methods, particularly when CV selection is unstable.

    Conclusions:

    • The proposed weighting method offers a more robust alternative to standard CV for NN-based tumor classification.
    • This approach enhances classification accuracy in the presence of uncertainty in parameter selection.
    • The findings suggest a potential improvement in diagnostic tools utilizing gene expression data.